Infrared Target Detection with Improved LBP Algorithm

Targets detection methods of infrared image sequences are studied in this paper. Based on analysis of the traditional LBP algorithm, an improved LBP algorithm, called ¿-LBP algorithm is proposed. This algorithm, which makes full use of the translational invariability and the rotational invariability of the LBP operator, is able to detect the edge of targets fast even under complicated backgrounds, in favor of targets recognizing and tracing. Comparing with several other kinds of traditional algorithms, this method improves the veracity and has fast responsivity.